The Google Whistleblower Taking on Big Tech Over Its Role in Gaza

By Evie Breese, UK Reporting Fellow at Tech Policy Press and openDemocracy | Fact-Checked by Irfan Ahmad

Image: Unsplash @Samtakespictures

The day Google DeepMind research engineer Sofia Liguori returned to work following the birth of her first child was the day her activism began.

The 32-year-old was part of the London-based frontier AI lab’s sustainability-focused Impact Accelerator team, tasked with identifying new ways the technology can be used for good. These range “from pathogen detection and crop resilience to medical countermeasure acceleration”, according to a LinkedIn post advertising new roles in the team.

Liguori’s own project focused on how AI could optimize electricity power flow to allow the national grid to use less energy and reduce overall carbon emissions. She told openDemocracy she now refers to her work as “building the nice face of the devil”.

“I felt so hypocritical that we were just talking about all these good things and not even mentioning Project Nimbus”, she said, speaking to openDemocracy from her home in south London, a Palestinian Keffiyeh scarf draped on the wall behind her.

Project Nimbus is a $1.2bn deal that Google and Amazon Web Services signed with Israel to build data centres in the country and provide public entities with cloud-based data storage in 2021. The contract stipulated that neither company could shut down their services in Israel or deny them to particular government entities.

Google has said its work with Israel is “not directed at highly sensitive or classified military workloads relevant to weapons or intelligence services”, as a spokesperson told Time magazine this year. Yet training materials leaked to The Intercept in 2022 revealed the full suite of Google Cloud AI tools are available to Israeli state departments, including the Ministry of Defense, for face detection, object tracking and “sentiment analysis.”

The ministry could have used these tools to identify and surveil Palestinians across the West Bank, East Jerusalem, and Gaza, including since 7 October 2023, according to reporting in the New York Times and the Center for Constitutional Rights. Israeli military officials have said they have killed 70,000 people in Gaza since the Hamas attack on October 7 2023 that killed approximately 1,200 Israelis.

As she sat at home with her new baby, Liguori said she “watched a live-streamed genocide for a year, knowing that my company was complicit”. She added: “If we accepted this modus operandi, we were basically legitimizing what has been tested in Gaza to become the standard.”

Returning to work in August last year, she said she implored her colleagues: “If our aim is AI for good, trying to stop genocide is one of the ultimate goods we can do, right?”

Ultimately, having been unable to convince executives to end Google’s role in Israeli military actions, Liguori left the company this year. In recent weeks, she has acted as a whistleblower, speaking out despite being subject to the strict lifetime non-disparagement agreement signed by Google employees.

Liguroi had always been prepared to be fired for her activism, but, she said, the real risk of “being sued by Google is different”. The tech giant’s parent company, Alphabet, is consistently ranked as the third or fourth most valuable globally, with a reported market worth of around $3.5trn.

This week, she was delivered some good news: a Google executive told the company’s UK staff that it will not enforce the NDA’s clause against workers who criticize the company after leaving. Speaking to openDemocracy after the news was announced, Liguori details what it took to get Google to back down.

The final straw

Liguori, whose background is in theoretical physics and reinforcement learning for video games, told openDemocracy she never set out to work with frontier AI models and knew “nothing” about Project Nimbus when she joined Google DeepMind in May 2023.

After becoming increasingly aware of the project’s impact in Gaza while on maternity leave, Liguori sent her immediate colleagues a letter urging them to take action shortly after she returned to work in August last year.

“This team is full of mothers,” she wrote. “I am not going to describe how it feels to become a mother for the first time while watching snipers targeting children, children dying of starvation, and all the rest of the horrors. What’s happening is lacerating my heart, really.”

The letter, which openDemocracy has seen, continues: “How can we do AI for good, and not even talk about Project Nimbus? How can we do AI for good in a company that is officially recognized and listed [as] complicit in the ongoing genocide in Palestine?”

Liguori also asked to discuss Nimbus in a team meeting. When her request was denied by a senior manager, she began sharing links to news articles about the project in internal channels. If they contained the word “genocide”, she said, they were deleted for “violating internal policies” prohibiting “aggressive language.”

Two months after her return to work, documents were leaked to Israeli-Palestinian publication +972 Magazine and Hebrew-language outlet Local Call revealing that Google and Amazon had agreed to use a secret code to “effectively sidestep legal obligations” and tip off the Israeli government if they were ever ordered to share Israeli data shared on their cloud services with law enforcement authorities anywhere in the world.

Then, in late April 2026, it was announced that Google had signed a contract with the US Department of War that would allow the Pentagon to use Google’s AI for “any lawful government purpose”, including mission planning and weapons targeting.

That was the day Liguori decided enough was enough. The petition she and over 600 colleagues had sent to CEO Sundar Pichai opposing the partnership had fallen on deaf ears.

“I really believe there is a value of infighting from the inside,” she said, “I tried to stay and do activism and raise awareness”, but the mental toll was becoming too great. “I reached a level of disgust with my company that was just incompatible with me continuing to stay there.”

Liguori used a combination of sick and holiday leave to cease undertaking work for Google, officially resigning on 20 October.

“I no longer believe that our technology will have a net positive impact on society. I think there’s a fantasy world where AI can help decarbonize grids, cut waste, protect ecosystems, and benefit humanity,” she wrote in her publicly posted resignation letter.

“But it’s not this world, and it’s not this type of general-purpose AI we are racing to develop.”

The growing revolt inside Google DeepMind

Google hasn’t always been so resistant to employee dissent.

Back in 2018, after 3,000 employees signed a petition seeking a policy that would stop Google and its contractors from “ever build[ing] warfare technology”, the tech firm announced that it would no longer be renewing its contract with the Pentagon on controversial military drone Project Maven.

Shortly after, the company published its AI Principles, a set of internal guidelines that prohibited the use of its technology for AI weapons or for human rights abuses.

But weeks after Donald Trump’s second term began in January last year, the Silicon Valley-based company appeared to have a change of heart, with the AI principles updated to remove references to weapons or surveillance.

“We believe democracies should lead in AI development, guided by core values like freedom, equality, and respect for human rights,” Google DeepMind chief Demis Hassabis and research labs senior vice president James Manyika wrote in a blog post announcing the updated policy.

Soon after, Google announced that Gemini would be used for both classified and unclassified purposes at the AI factory of Lockheed Martin, a US weapons company that primarily builds military aircraft, missiles, satellites, and defense technology for the US armed forces.

Then, in June of this year, a partnership between Google and US military contractor Palantir was expanded to integrate Palantir platforms across Google Cloud and Google Cloud Marketplace.

“You never know then how your technology will be used in Palantir’s systems,” Liguori told openDemocracy. “Workers need to know what the technology they are building on will contribute, and we want the right to abstain on ethical grounds”.

Since then, a number of UK Googlers have left the company on ethical grounds, but were gagged from speaking about the firm by strict NDAs. Feeling strongly that the public needed to know about Google’s dealings with Israel, Liguori said she decided to breach her own NDA to publish her resignation letter, knowing that Google could retaliate.

But this week Google told its staff in the UK that it will not take legal action against those who criticize the company or its AI products after leaving. “We’ve always encouraged an open dialogue about safety and responsibility,” a Google spokesperson told The Washington Post. “Recently, we’ve updated our terms for new hires, which included removing the post-termination aspect of non-disparagement clauses.”

Liguori sees the move as a win that “happened because of collective action”. After an employee’s attempts to raise concerns about the clause through internal channels did not lead to any change, DeepMind members of the United Tech and Allied Workers (UTAW) coordinated with the No Tech for Apartheid campaign group and whistleblowing non-profit Psst to pressure Google from multiple directions, Liguori told openDemocracy.

It’s only because of this support network that Liguori says she was able to “find the courage to take the risk” of a public resignation.

The concession comes as the Employment Rights Act 2025 comes into effect in the UK, allowing unions to enter workplaces across the country to speak to workers in person about joining a union. At Google and Google DeepMind, the new law will likely support UTAW and Unite’s drive for formal recognition.

For Liguori, who joked that she had considered “disappearing in the Italian countryside”, the news that she won’t face legal action for her activism has come as a “big relief”. She added: “... it’s also a sign that our collective effort had an effect, and that it's worth challenging the rules.”

Google has not responded to openDemocracy’s request for comment at the time of publication.

Read next:

• Hate Speech Still Pays on YouTube. Regulators Should Follow the Money

• How to Navigate Difficult Family Relationships During Holiday Gatherings

by External Contributor via Digital Information World

How to Navigate Difficult Family Relationships During Holiday Gatherings

By Marina Zhukova, UTHealth Houston | Reviewed by Irfan Ahmad

Image: Rein Krijgsman - Unsplash

Family gatherings can bring joy, togetherness, and celebration, but they can also bring unexpected stress and unresolved conflicts. Conflicts are normal in any relationship, but they don’t have to overshadow the holiday.

What does unhealthy family conflict look like? How can you differentiate it from typical family stress?

Disagreements are a normal part of any relationship, including relationships with family members. In fact, we may react more strongly to disagreements with loved ones because we often expect them to share our values and beliefs. There is a growing trend on social media to label relationships that involve conflict as “toxic.” However, experiencing conflict does not necessarily mean that a relationship is unhealthy. Disagreements are inevitable, and maintaining meaningful relationships requires effort. We can care deeply about someone while holding very different opinions or making different life choices.

The key distinction is not whether conflict occurs, but how it unfolds. Typical family stress may involve occasional arguments or frustration, but family members can generally move past disagreements while maintaining mutual respect. Unhealthy conflict, on the other hand, tends to involve recurring patterns of criticism or hostility, particularly when one person’s feelings or needs are consistently dismissed. It may leave someone feeling afraid to express themselves or constantly anticipating the next confrontation. Rather than focusing on a single unpleasant interaction, I encourage people to consider the overall pattern of the relationship.

Why does family conflict intensify during the holiday season?

Three main things come to mind when thinking about family conflict around the holidays:

First, we’re spending time with relatives we don’t see regularly. When we don't interact with someone often, we may be less familiar with how to communicate effectively with them or navigate differences in opinions.

The holidays can also mean returning to familiar environments. In psychology, we know that environmental cues can trigger learned patterns of behavior. For example, returning to your childhood home may bring back old family dynamics. You might find yourself responding to your parents or siblings much like you did as a teenager, even though you are now an adult. Parents may fall back into their old roles, as well.

This time of year also brings increased stress and changes in routine. The holidays often involve travel, disrupted routines, and additional responsibilities. All of this can leave us with fewer resources to regulate emotions and, in turn, to manage disagreements.

What are some strategies for coping with family conflict?

• Cope ahead: Anticipate difficult situations and plan how to respond. For example, if you expect uncomfortable questions about your career, marriage, or plans to have children, prepare a few responses. You might start with humor, redirect the conversation, or politely say you would rather not discuss it. If the topic has come up many times before, it is okay to remind the person that your perspective has not changed.

• Maintain routines that help you manage stress: Even during busy family visits, try to make time for activities you enjoy, whether that is a morning walk, meditation, or listening to your favorite podcast. Taking breaks from family interactions can also help you recharge.

• Plan some downtime before and after family gatherings: If possible, avoid rushing straight from a stressful visit back into work or other responsibilities. Giving yourself time to decompress can make the experience more manageable.

What are some healthy ways to de-escalate conflict?

• Pick your battles: Not every disagreement needs to be resolved, especially during the holidays. Sometimes it is better to acknowledge that you have different opinions and move on. Humor can be a great way to diffuse tension and redirect the conversation, as long as it is not at someone else’s expense.

• Validate the other person’s feelings without necessarily agreeing with them: For example, saying, “I know this is important to you,” can help someone feel heard. At the same time, be direct about your own feelings. In dialectical behavior therapy, we teach communication skills that emphasize describing the situation without judgment and clearly expressing what you need. Instead of saying, “You always criticize my choices,” try, “When you bring up my career decisions, I feel pressured. I'd appreciate it if we could talk about something else.”

• Know when to take a break: If you notice yourself becoming frustrated or raising your voice, step away from the conversation. Even a short pause can help you regulate and respond more thoughtfully. You do not need to resolve years of family disagreements over one holiday dinner.

How can someone set boundaries without escalating conflict?

It is important to remember that healthy boundaries are about managing our own behavior, not controlling someone else’s. There is an old saying that “our freedom ends where another person’s nose begins.” I think this applies well to boundaries. We have the right to decide what we are comfortable with, but we also need to respect other people’s autonomy.

Be clear and specific about what you need, without blaming the other person. For example, instead of saying, “You need to stop asking about my personal life,” you might say, “I know you care about me, but I’m not comfortable discussing this. I’d rather talk about something else.”

Setting a boundary does not guarantee that the other person will agree with it. You can communicate your limits respectfully, but you cannot control their reaction. If someone continues to push, it is okay to calmly end the conversation or step away.

What advice do you have for someone navigating a difficult family member?

The approach would depend on how much you value the relationship and what you hope to get out of the interaction.

If this is someone you rarely see and don’t have a particularly close relationship with, it may be best to focus on keeping the interaction manageable. Be polite, pick your battles, and remind yourself that the gathering is temporary.

If the relationship is important to you, it may be worth investing more effort into improving communication. We call this interpersonal effectiveness. Approach the person with curiosity rather than judgment, try to understand their perspective, and acknowledge their feelings without necessarily agreeing with them. Be open about your own feelings, as well. Remember that improving a relationship takes effort from both sides, and you can do your part without taking responsibility for the other person’s behavior.

Do not put additional pressure on yourself to have a perfect holiday. It is okay to experience mixed emotions, including frustration or disappointment, while still enjoying time with your family.

By: Marina Zhukova, PhD, psychologist and assistant professor, Louis A. Faillace, MD, Department of Psychiatry and Behavioral Sciences, McGovern Medical School at UTHealth Houston. All quotes should be attributed to her.

Read next: Hate Speech Still Pays on YouTube. Regulators Should Follow the Money
by External Contributor via Digital Information World

Hate Speech Still Pays on YouTube. Regulators Should Follow the Money

By Richard Woods, Tech Policy Press | Fact-Checked by Irfan Ahmad

Image: Sora Shimazaki - Pexels

For every dollar an advertiser spends on a YouTube video, the platform keeps 45 cents. YouTube is not a neutral host that occasionally misses bad actors. It is a business partner in every monetized video, and it profits when those videos perform, including, new research suggests, ones that could break its own rules.

A new report from the Global Disinformation Index (GDI), “Monetizing Hate,” puts a price on that partnership. GDI estimates that 100 English-language channels publishing what it classifies as hate speech are on track to earn about $13.3 million in ad revenue in 2026: $7.3 million for creators, $6 million for YouTube. One channel, earning an estimated $2.2 million a year, carried content that may breach YouTube’s Community Guidelines and German law. In September, it was still online, still showing ads and still viewable in Germany.

The report’s most revealing finding is not the size of these sums. It is how little anyone outside YouTube can know about them, and how conveniently that serves the company.

Little money, strong incentives

Six million dollars is a rounding error for a platform that took in $60 billion from ads and subscriptions in 2025 alone. The more plausible explanation for weak enforcement is not that hate is too lucrative to give up, but that enforcement costs time, staff and creator goodwill, while inaction costs little.

For creators, though, $2.2 million a year is a career. Recommendation systems tend to reward content that edges toward a platform’s policy line, and because YouTube shares the revenue, its interest in a high-performing video is aligned with the creator’s, not with the people that video targets.

GDI’s numbers deserve scrutiny. Five large language models labeled 19,297 videos; analysts manually reviewed 200. Its definition of hate speech follows the Council of Europe’s and is broader than YouTube’s, so much of what it flagged is permitted on the platform. Critics will say GDI is counting lawful speech. But that gap is the point: YouTube’s narrow definitions decide what is monetizable, and only YouTube can test GDI’s estimates against real data.

YouTube’s repeat-offender clause applies in “rare cases” and never defines “repeatedly.” GDI found no evidence it had been used against the content it reviewed. A rule with an undefined trigger and unreported use is not really a rule. It is discretion. YouTube pledged to tighten its rules on monetizing hateful content after an advertiser revolt in 2017 and again in 2019. The pattern suggests enforcement tightens when revenue is threatened, then slackens when attention moves on.

What meaningful transparency would look like

YouTube’s transparency reports count removals, not videos that stay online but lose ads. Meaningful transparency would measure money, not takedowns. In practice, that means:

• Enforcement data that matches the business model: how many videos and channels are demonetized under the hate speech policy, by country of viewing.
• A defined repeat-offender threshold, with annual figures on how often it is triggered. A rule YouTube must report on is a rule it must use.
• A measure of delay: how much a video earned before demonetization, and how long it took. This shows whether enforcement removes the incentive or arrives after the money is made.
• Video-level placement reports, so advertisers can see, and stop paying for, what their budgets fund.

Much of this sits within existing law. The Digital Services Act (DSA) requires platforms to explain any “restriction of monetary payments” to the affected user (Article 17), and those explanations feed the Commission’s public Transparency Database. Regulators could check whether YouTube logs hate speech demonetizations there at all. The Commission can demand data on viewer location and revenue under Article 67, and a delegated act adopted in July 2025 lets vetted researchers request non-public platform data. In the UK, Category 1 transparency reports under the Online Safety Act, due in 2027, give Ofcom a chance to require demonetization figures, not just removals.

Disclosure has limits. YouTube’s own DSA audit, as GDI summarizes it, found trusted-flagger notices alleging illegal hate speech were not prioritized. Reporting can harden into box-ticking, and a demonetization figure means little without an independent sense of how much violating content exists. That is why researcher access matters as much as company reporting.

Following the money sidesteps the censorship fight

US-based creators account for 81% of the estimated revenue, an easy target in the transatlantic row over the DSA. But EU users were shown ads on sampled videos. If even 10% of views came from the EU, GDI estimates, that would mean $0.9 million in hate-linked ad revenue from January to August. This is only an estimate because YouTube alone knows where its viewers are.

Demonetization removes no video and silences no one. It asks a company to stop paying for, and profiting from, content that breaks its own rules or the law where it is viewed. That shifts the question from “what should come down” to “what should be paid for.” A platform that cannot answer it with data has already told regulators something.

This story was originally published by Tech Policy Press.

Read next: Can ChatGPT Catch Up With Google Ads? Study Reveals a Significant Advertising Performance Gap

by External Contributor via Digital Information World

Can ChatGPT Catch Up With Google Ads? Study Reveals a Significant Advertising Performance Gap

By Frankfurt School of Finance & Management | Reviewed by Irfan Ahmad

Advertising on ChatGPT has so far proven significantly less cost-effective for businesses than advertising on Google. This is the finding of a new study by Christian Schulze, Associate Professor of Marketing at Frankfurt School of Finance & Management, and Maximilian Kaiser, Assistant Professor at Vienna University of Economics and Business (WU Vienna). The researchers examined the effectiveness of paid ChatGPT ads and their impact on user behavior. The findings raise questions about OpenAI’s business model and its potential initial public offering (IPO).

ChatGPT advertising significantly less effective than Google Ads

While a visit generated through a Google search ad produced an average of $2.45 in revenue, a visit from a ChatGPT ad generated just $0.89 – approximately 64 per cent less. Conversion rates were also considerably lower, at 0.37 per cent for ChatGPT ads compared with 1.17 per cent for Google search ads. Among the 102 advertisers with complete revenue data, 82 per cent did not record a single purchase from their ChatGPT ads. Overall, at publicly reported ChatGPT ad prices, Google search advertising delivered 6.6 times better value for money in terms of revenue generated per advertising dollar.

Image: eCommerceResearch

“ChatGPT is still a long way from becoming an economically viable alternative to Google as an advertising platform. What matters to businesses is whether advertising leads to purchases and delivers a return on investment. This is precisely where we see substantial weaknesses. Remarkably, even ChatGPT’s unpaid recommendations perform substantially better than its paid ads,” explains Christian Schulze.

More advertising could undermine ChatGPT’s appeal

The study also suggests that, following the introduction of advertising, ChatGPT referred approximately six to eight per cent fewer visitors to online shops through unpaid recommendations. However, the conversion rates of visitors who continued to arrive through these recommendations remained unchanged.

This presents OpenAI with a dilemma: To increase advertising revenue, the company could integrate ads more closely into ChatGPT’s responses. However, doing so could undermine users’ trust in the independence of its recommendations.

“OpenAI faces a fundamental trade-off: Advertising must be commercially attractive without compromising the value of ChatGPT for its users. Integrating ads more closely into responses could improve their effectiveness, but it could also undermine user trust,” says Christian Schulze.

Study raises questions about OpenAI’s growth plans

The findings are also relevant in the context of a potential OpenAI IPO. According to company figures cited in the study, providing free access to ChatGPT costs approximately $5 per user annually, while projected advertising revenue currently amounts to only around $1 per user. To cover these costs and achieve its ambitious revenue targets, OpenAI would need to significantly increase the economic value of its advertising. The key challenge will be to achieve this without compromising ChatGPT’s appeal to users.

About the study

For their study, “Can Advertising Fund Free AI? Evidence from ChatGPT’s International Advertising Rollout”, the researchers analyzed data from 1,898 international online shops, covering nearly ten billion website visits between June 2025 and August 2026. Advertising effectiveness was assessed using data from 387 websites, including 102 with complete revenue data. Detailed information on the study is available for free at ecommerceresearch.org/chatgpt-ads.

Read next: How to prepare your online life for your death

by External Contributor via Digital Information World

How to prepare your online life for your death

By Jenny Tang, Carnegie Mellon University and Xiaoyuan (Owen) Wu, Carnegie Mellon University | Fact checked by Irfan Ahmad

Image: Kampus Production - pexels

What happens to all of your online accounts – credit cards, merchants, news subscriptions, social media, streaming services – after you die? How will your heirs know how to access these accounts or understand what you want done with them? Most people haven’t considered these questions, and this lack of planning can create significant headaches for next of kin.

We studied challenges older adults encounter when thinking about what should happen to their online accounts after they die. We interviewed 21 American adults over age 60 to learn about their concerns, needs and practices for their online accounts, as well as what would make planning for these accounts easier.

Most of the people we interviewed had hired lawyers or estate planners to draft wills specifying how their money and physical property should be handled after their deaths. They wanted to make sure that heirs had easy access to the money and assets left behind.

People also wanted their relatives to have easy access to their digital estates, but they had not done any formal planning and didn’t know how to plan for transitioning or shutting down online accounts, such as email, social media and accounts for online bills. Some told us they hadn’t thought about their online accounts at all. Others said they assumed heirs would know what to do, or they assumed that giving heirs the passwords to accounts would be enough to grant access to their digital estates.

Unfortunately, more planning is typically necessary to smoothly transition a digital estate.

Passing down a list of passwords

The older adults we interviewed mainly kept track of their passwords by hand: They wrote their passwords down on paper, or they stored them in a digital file. Almost everyone had already shared some of their passwords with family members who help them manage their accounts. Everyone also believed that their heirs would be able to find their list of passwords after their death.

The people we interviewed weren’t always happy with their current systems for keeping track of their passwords. Some people complained that when accounts required frequent password changes, it was difficult to keep their password lists up to date. As a result, they weren’t sure their heirs would get the latest version of all their passwords.

Password managers can be a good way to make sure passwords are up to date and, if set up appropriately, accessible to a designated heir. While most of the people we interviewed used password managers, some said they didn’t save important passwords in them because they didn’t know whether they were secure. However, using a password manager is the most frequent security recommendation from experts, who typically use them to store passwords for all of their accounts.

By using a password manager, you can make sure your designated heir has access to all your latest passwords. Some password managers provide access for survivors by default, such as through time-delayed emergency access features. You can also write the master password down somewhere safe to pass to your heir. Instead of making sure dozens of account passwords are up to date, you only have to manage one.

Passwords alone cannot unlock all accounts

Unfortunately, just handing someone your passwords doesn’t always mean they can log in and manage your account. If the account asks for two-factor authentication, such as through a code sent to a phone or email, a password alone won’t grant access. And if the email account is closed or the phone number is no longer active after you die, then they might be out of luck.

To make sure your heirs aren’t locked out, identify the devices and accounts that can act as keys for heirs to unlock the rest of your online accounts, and direct your heirs to keep them open until other accounts have been settled. This not only makes sure that heirs can get authentication codes, but also lets heirs reset passwords to your other accounts if they can’t find the password.

In some cases, passwords might not be necessary. Some online services, such as Facebook and Google, have features that let you choose someone to manage your account or get your data after you die. However, not all platforms offer this option, and few of the older adults we interviewed were aware of these features.

Online accounts don’t automatically die with you

Like many of the people we interviewed, you might also assume that heirs will properly deal with accounts with money and important data, while unimportant accounts can be ignored. But even if there is nothing important in an account, it can still be used by hackers for identity theft, spam or fraudulent activities. Automatic monthly charges, to cover the cost of streaming services or annual billing for magazine subscriptions, for example, might also continue for a long time.

Without instructions, the people managing your online accounts after you die might not think to close those accounts. They might also find it hard to guess your intentions.

You can leave instructions for your heirs by going through your accounts and noting which ones can be closed outright, which ones require final payments before closure, and which ones might require other actions. You can also periodically tidy up and close accounts you are no longer using to keep your list of accounts more manageable.

Planning ahead for you and your heirs

While it seems simple to just let your heirs handle your important online accounts and let the others slowly fade into nonexistence after you die, the reality is often more complex.

Your online accounts will live on after you die. Planning ahead for what happens to those accounts can reduce the stress on your heirs and ensure your digital accounts are handled according to your wishes.

Lujo Bauer, Nicolas Christin and Lorrie Cranor contributed to this article.

Jenny Tang, Ph.D. Student in Societal Computing, Carnegie Mellon University and Xiaoyuan (Owen) Wu, Ph.D. Student in Societal Computing, Carnegie Mellon University

This article is republished from The Conversation under a Creative Commons license. Read the original article.

Read next: 72% of students say they procrastinate. Here’s how to stay on track

by External Contributor via Digital Information World

63% Followed Unverified AI Health Advice, 58% of Millennial Parents Did So for Children, 42% Misrepresented Information for Privacy Reasons

By Talkerresearch | Reviewed by Irfan Ahmad

One in seven Americans trust AI chatbots as much as they trust a doctor, according to new research.

A survey of 1,000 millennial parents and 1,000 general population Americans explored how they are using AI to manage their own health or their children’s health, and whether they understand how AI generates the advice it provides.

Image: Betterimagesofai - CC BY

Confidence in AI is growing

Among Americans who use AI for health purposes, the number of respondents who trust AI as much as they trust a doctor nearly quintuples compared to non-AI users, and 60% said they would believe AI-generated health advice even if it contradicted guidance from a healthcare provider.

The survey conducted by Talker Research for MD Live by Evernorth found that nearly two-thirds (63%) of AI users admit they have followed health advice from AI without first confirming it with a healthcare provider.

That confidence may stem from the way AI delivers information: Half of respondents said that if AI gives them a clear or reassuring response about their health, they’re likely to accept it without questioning it.

Those who use AI for health purposes ask an average of nine health-related questions each week.

These respondents say they rely on it because it’s conveniently available at any time (37%), faster than seeing a doctor (31%) and provides reassurance about their health concerns (30%).

However, 28% of respondents say they only verify the answer they receive from AI less than half the time.

The trend extends to decisions about children

Among millennial parents who use AI for health purposes, 58% admit they have acted on AI-generated health advice for their children without first confirming it with a healthcare provider.

Many also said they would be comfortable solely relying on AI for a range of decisions involving their children’s health, including evaluating minor symptoms (80%), determining what type of care is needed (71%) and seeking medication guidance (71%).

Despite the confidence in AI, Americans often overlook how their input may affect the accuracy of the guidance it provides. Nearly eight in ten respondents (78%) did not recognize that the way a question is worded influences the answer AI provides.

More than half (53%) admitted they have reworded or adjusted their health question when they didn’t like the first answer they received, including 27% who say they always do so.

Many AI users also ask questions in ways that steer AI toward a desired response.

Thirty-eight percent said they describe symptoms and ask what could be wrong, while about a third ask whether a symptom could be connected to a condition they already suspect they have (34%) or whether a specific diagnosis could explain what they’re experiencing (32%).

“The survey indicates that people are often going to AI already hoping for a certain answer, whether that’s reassurance or confirmation of a fear,” said Dr. Maggie Williams, medical director of virtual primary care at MD Live by Evernorth. “That’s very different from how a doctor works through a symptom with a patient, asking questions, gathering context and ruling possibilities in or out before reaching a conclusion.”

Many overestimate what AI actually knows

The survey also found that many Americans overestimate how AI works in the first place.

One in five respondents believe that AI only pulls information from trusted medical sources (22%), and the same share of regular users believe AI factors in their personal medical history even when that information has never been provided.

At the same time, many respondents acknowledge they do not provide AI with a complete picture of their health when seeking guidance.

Forty-two percent admit they’ve even lied about who they’re submitting a health question about for privacy reasons.

More than three-quarters of regular users (76%) leave out their medical history, demographic details (76%) and personal information due to privacy concerns (81%).

As a result, many AI users report receiving answers that weren’t fully relevant to their circumstances.

Nearly six in 10 (59%) said they later realized health information they received from AI didn’t fully apply to their situation, and nearly half (49%) say they acted on that information anyway before discovering it wasn’t the right fit.

Human oversight remains important

Despite growing adoption of AI for health purposes, respondents expressed concerns about receiving incorrect advice (34%) and a lack of human oversight (31%).

One in four said they plan to use AI for more healthcare decisions in 2027, while 27% said they would trust AI-generated information more if it is reviewed by a doctor.

“The results show people will continue to turn to AI for their health care questions — and having people more engaged in their health is a good thing,” said Dr. Williams. “But healthcare providers add important clinical guidance by catching details a patient may not consider relevant, asking follow-up questions and putting the information into the context of that person’s health history. As we consider a patient’s full care journey, it’s important to understand how AI and physicians can come together to provide the clearest picture to patients while ensuring appropriate safeguards and clinical oversight.”

Read next: Study Finds 80% Missed Chatbot Errors Despite Participants’ Basic Knowledge of the Subject

by External Contributor via Digital Information World

Study Finds 80% Missed Chatbot Errors Despite Participants’ Basic Knowledge of the Subject

By Radboud University | Reviewed by Irfan Ahmad

A chatbot that speaks naturally is more likely to be perceived by users as intelligent and useful, even if it spouts nonsense, research by communication scientists at Radboud University shows. An eloquent AI system that shares incorrect information is rated just as highly by users as a curt AI system that provides correct answers.

Image: Solen Feyissa - Unsplash

Most people are by now quite familiar with ChatGPT, Gemini, Claude and other chatbots: every day, people ask millions of questions and receive neatly formulated answers in return. Yet these language models still tend to run wild with so-called ‘hallucinations’: unreliable, dubious answers that are presented with great certainty as the truth. Communication scientist Marianna de Sá Siqueira and her colleagues investigated the effects of elegant language use on the credibility of chatbots.

Nonsense in fluent language

For the experiment, participants were shown one of four variants of a conversation with a chatbot about The Lord of the Rings. Two variants featured fluent sentences and human-like language, one of which gave correct answers and the other did not. In addition, there were two variants with short, businesslike answers, one of which provided entirely correct information whilst the other presented (subtly) fabricated facts.

Unsurprisingly, users were most positive about the chatbot with elegant language and correct answers. However, a chatbot that writes fluently but spouts factual nonsense was rated by participants as just as intelligent and pleasant as a chatbot that communicates in a stilted manner but does provide factually correct information. It also emerged that as many as 80 per cent of participants did not notice the errors at all, even though they had been selected in advance for their basic knowledge of The Lord of the Rings.

"Above all, it shows how easily people can be led astray and how quickly we come to trust something that is actually spouting nonsense," says Barbara Müller, associate professor and co-author of the study. "If a text sounds elegant and natural, people automatically attribute intelligence to it. You then subconsciously tend to think: this is fine, I don’t need to check it any further."

Blind trust

The research highlights the danger of blind trust in artificial intelligence. Professor and co-author Tibor Bosse says: "So-called language models such as GPT and Claude are primarily trained to produce statistically plausible and easy-to-read sentences. These systems do not always take the truth very seriously. They are designed to sound human and convincing, but whether the information is factually correct is of secondary importance."

"People used to complain that chatbots were annoying and knew nothing. Now they sound so clever that we’re too quick to believe them," says Bosse. According to him, this cannot be solved with a simple disclaimer at the outset. "Most chatbots show a single line of text at the start stating that the output may not be entirely accurate, but those companies are certainly not going to indicate of their own accord with every answer that the system is uncertain. If we really want to protect users from chatbots spouting nonsense, we need to, for example, oblige tech companies to make that uncertainty visible. If a chatbot itself expresses doubt, you prevent people from taking nonsense as the truth."

Literature reference:

de Sá Siqueira, M. A., Bosse, T., & Müller, B. C. N. (2026). Users’ perception of chatbot language fluency and hallucination. Computers in Human Behavior: Artificial Humans, 10, 100412. https://doi.org/10.1016/j.chbah.2026.100412.

Read next: Deceptive Google Ads Generated Over 100 Million Impressions Despite Existing Safeguards, Some Running for Years

by External Contributor via Digital Information World